a month ago
Hi! I'm decommissioning an old project and hit something odd: the project/service is deleted, but one of its containers is apparently still running and publicly reachable.
What happened:
- 2026-07-14: got the "Upcoming Project Deletion" email (project queued by our team; email said active deploys were stopped, permanent deletion within 48 hours).
- 2026-07-16 ~15:40 UTC: right at the end of that 48h window, a container from the project (re)started - I can tell from its /health endpoint's uptime counter.
- 2026-07-16 17:00 UTC: I also deleted the project via the CLI (
railway project delete); it was accepted and deletedAt was stamped. My workspace shows no projects now. - Today (2026-07-21): that container is still up. /health returns 200 with a continuous uptime of ~5 days, spanning the deletion. Its *.up.railway.app subdomain 404s, but a custom domain that was attached to the service still routes to it through the Railway edge (server: railway-hikari).
What I'm hoping for:
- Hard-deletion of any containers/services still running under that project ID.
- Release of the custom-domain claim so the edge stops routing that hostname.
I'm was on the hobby plan so I can't open a private ticket. Happy to share details with staff via DM. Is this something support can pick up from here, or is there a better channel for account-specific cleanup on a free plan?
P.S.: have no service to attach it to because everything is deleted UI-side.
2 Replies
a month ago
Your account shows no remaining services or domains, so both the container and domain claim should have been cleaned up with the project. Could you share the custom domain that is still routing traffic? That will let us locate the orphaned resources.
Status changed to Awaiting User Response Railway • about 1 month ago
Railway
Your account shows no remaining services or domains, so both the container and domain claim should have been cleaned up with the project. Could you share the custom domain that is still routing traffic? That will let us locate the orphaned resources.
a month ago
It still returns HTTP 200 from https://api.helloamicus.com/health as of 2026-07-21 18:43 UTC, through the Railway edge (server: railway-hikari, x-railway-edge: jfk1). A request ID from that probe, in case it helps you trace it: x-railway-request-id: goSCOBooRzutUgFxn6XIxQ
A few details that may help locate the orphaned resources:
- The container's /health uptime counter puts its start at 2026-07-16 15:40:41 UTC, about 80 minutes before the project's
deletedAttimestamp and it has run continuously since, suggesting it survived the deletion rather than being restarted after it. - The service's public subdomain was klqbs2vk.up.railway.app; that now 404s, so the only surviving route is the custom-domain claim.
- The domain was attached to service
@amicus/server(e1d975e1-1d50-4175-b747-ae29ef790f9b), production environment (d9ee58e2-0a7f-4c67-9dff-dd5bdd613652) of project5417afd8-59b5-440d-b7ca-bc3524ff3c48.
Since the project is deleted, could you destroy the remaining container and release the custom-domain claim so the edge stops routing that hostname? Happy to share anything else that helps via DM. Thanks
Status changed to Awaiting Railway Response Railway • about 1 month ago
a month ago
This thread has been opened as a bounty so the community can help solve it.
Status changed to Open Railway • about 1 month ago
Status changed to Solved hgimonet • 30 days ago